Oracle FS1-2 Flash Storage System
 

PLA-Support>Sun Systems>DISK>Flash Storage>SN-EStor: FSx
 

In this Document

Symptoms

Changes

Cause

Solution

References

Created from <SR 3-13445672841>

Applies to:

Symptoms

 There is a System Alert for "Volumes Require Migration".

Changes

The FS1 System Software version was upgraded from 6.1.x to 6.2.x

Cause

This is intended behavior.

Beginning with Version 6.2.0, the Oracle FS System software prevents the creation of volumes with single parity (RAID 5) on the Capacity Disk Storage Class. All new volumes are created with double parity (RAID 6) on the Capacity Disk Storage Class for added data protection in the event of a second drive failure in the same Drive Group. Existing volumes that were created with single parity (RAID 5) are not immediately affected but if these single parity volumes undergo maintenance, such as being unmapped and then remapped, the system will automatically migrate them to double parity.
Systems that created RAID 5 type parity Volumes on the Capacity Disk Storage Class on releases below 6.2 will generate a system alert on a software update or system startup to indicate that there are existing RAID 5 Volumes that need to be migrated to RAID 6.

 

Solution

The RAID 5 volumes on the Capacity Disk Storage Class should be migrated to RAID 6 at your convenience, one at a time, keeping in mind that each volume migration will cause somewhat of a performance impact.

Multiple migrations should not be done at the same time unless the Volumes are small and the system load is light enough that the performance impact of the migrations will not be an issue.

To determine from the GUI which LUNs require the modification from RAID 5 to RAID 6:

1. Navigate to the SAN tab and click on "LUNs" under "Storage".
   
2. In the right window, under the "Tier Reallocation" column, click on any LUN that has a double dash (--) in that column. This denotes a LUN that was created using the Single Tier option (not Auto-Tiered). Then click on the "Tier Reallocation" column heading to sort the column and group the Single Tier LUNs to make it easier to find all of them.
3. Right-click on each Single Tier LUN, one at a time, and select View.
4. On the Quality of Service tab, check the Storage Class and the RAID Level. If the LUN's Storage Class is Capacity Disk and it's RAID level is Single Parity (RAID 5), then that LUN requires modification to Double Parity (RAID 6).

To migrate the LUN to RAID 6, right-click on the LUN name and select Modify. On the Quality of Service tab, change the RAID Level to Double Parity. If you cannot click on the drop-down arrow to change it (because it's "grayed out"), then un-check the "Use Storage Profile" check box first. Click OK and the migration will start. Note that any Quality of Service LUN migration will create somewhat of a performance impact.

The System Alert will return if the pilot is failed over to the standby pilot or if the system is restarted and there are still LUNs that have not been migrated.

If there are a very large number of LUNs, clicking through all of them may not be practical. In that case you can open a Service Request, upload a current manual log bundle, and a TSC engineer can generate a list of the LUNs that need migration.

Despite what it says in the software Release Notes, a System Alert will not be generated for each LUN that requires migration. Only one System Alert is generated, and it does not name the LUNs. See Bug 23612798 for more information.

There is not an easy way for the customer to determine from the GUI or fscli which LUNs require modification.

TSC Engineers can identify the volumes in question by running the following commands on a log bundle that has already been scanlogged:

The numbers in the awk command above refer to the column headings from the FSInfo.pl -l output, and these are the columns that will be returned by the awk command:

1 = LUN Name

12 = Storage Class

19 = Slat Enabled

25 = Raid Configuration

Example:

-bash-4.1$ awk -F '|' '{print $1 $12 $19 $25}' lunlist.out | grep STORAGE_CLASS_NEARLINE_HDD | grep RAID_5
VMWFS1_LUN069_TEST_SQL      STORAGE_CLASS_NEARLINE_HDD      false      ReadAhead: NORMAL, Raid Level: RAID_5
VMWFS1_LUN155_PROD_SQL      STORAGE_CLASS_NEARLINE_HDD      false      ReadAhead: NORMAL, Raid Level: RAID_5
VMWFS1_LUN169_PROD_SQL      STORAGE_CLASS_NEARLINE_HDD      false      ReadAhead: NORMAL, Raid Level: RAID_5
INCORPSERV207_R                     STORAGE_CLASS_NEARLINE_HDD      false      ReadAhead: NORMAL, Raid Level: RAID_5
-bash-4.1$

Alternatively, you can open the *.chsh.xml file and search for GetSystemAlertResponse:

<GetSystemAlertResponse>
<SystemAlertInformation>
<Identity>
<Id>414B303032363535A13E0062B7152A71</Id>
<Fqn>/VolumesRequireMigration</Fqn>
</Identity>
<CreationDate>2016-09-16T02:16:14.603</CreationDate>
<VolumesRequireMigration>
<VolumeIdentity>
<Id>414B303032363535A104AE2DFAA4DDD4</Id>
<Fqn>/VMWFS1_LUN069_TEST_SQL</Fqn>
</VolumeIdentity>
<VolumeIdentity>
<Id>414B303032363535A104E28E22360BD0</Id>
<Fqn>/VMWFS1_LUN155_PROD_SQL</Fqn>
</VolumeIdentity>
<VolumeIdentity>
<Id>414B303032363535A10471C10303E506</Id>
<Fqn>/VMWFS1_LUN169_PROD_SQL</Fqn>
</VolumeIdentity>
<VolumeIdentity>
<Id>414B303032363535A104A83602949C64</Id>
<Fqn>/INCORPSERV207_R</Fqn>
</VolumeIdentity>
</VolumesRequireMigration>
</SystemAlertInformation>
</GetSystemAlertResponse>

References

This solution has no attachment